This document discusses modeling industrial demand response in wastewater treatment. It summarizes that wastewater treatment is electricity intensive in Europe. Case studies show potential for flexible operation of equipment like aeration and pumping for up to 120 minutes. However, quantifying demand response potential for the entire wastewater sector has not been done. The document proposes an integrated energy systems model that combines a power system model with a three-component wastewater treatment process model to more accurately represent demand response while considering process constraints.

1. Modelling industrial demand response:
The case of wastewater treatment
DATE
24/09/2020
AUTHORS
Dana Kirchem
Muireann Lynch
(The Economic and Social Research
Institute)
Matteo Giberti
Recep Kaan Dereli
Eoin Casey
(University College Dublin)
Juha Kiviluoma
(VTT Technical Research Centre of
Finland LTD)

2. /12
Dana Kirchem dana.kirchem@esri.ie
2
Electricity consumption in the water sector by process and region in 2014 (IEA, 2016).
Electricityconsumptionin
consumption
Wastewater treatment in Europe is an electricity-intensive industry.

3. /12
Dana Kirchem dana.kirchem@esri.ie
3
Case studies find potential for flexible operation of wastewater treatment equipment.
Aeration
Pumping
Biogas
Can be turned off for between 15 minutes (Berger et al., 2013) and
120 minutes (Schaefer et al., 2017).
Switch-off times up to 30 minutes for inlet pumps (Nowak et al., 2015) and 60
minutes for recirculation pumps (Schaefer et al., 2017)
Flexibility potential arises from biogas storage and use for on-site
electricity generation.

4. /12
Dana Kirchem dana.kirchem@esri.ie
4
There has not yet been an attempt to quantify
the demand response potential of the
wastewater treatment sector on an energy
system level.
However ...

5. /12
Dana Kirchem dana.kirchem@esri.ie
5
Traditional power system models often model demand response as a virtual storage.
Thermal power
generation
Fossil fuel supply
Renewable power
generation
Renewable input (e.g.
Weather profiles)
Non-responsive
demand
Flexibility options (Hydro, Batteries, Power-to-X…)
Supply Demand
Model input
Responsive
demand
(Virtual storage)
Energy Systems Model
Objective function:
• Cost minimisation at system level
Optimisation variables:
• Level of capacity dispatch
Constraints/parameters:
• Coverage of non-responsive demand
• Technical constraints
• Policy constraints

6. /12
Dana Kirchem dana.kirchem@esri.ie
6
This approach is insufficient for many industrial demands for two reasons.
1. It neglects process constraints within the industrial process and distort results.
 Intertemporal interdependency of production processes and the precision in timing
 Input and output constraints (e.g. wastewater inflow and effluent quality)
2. Reliable energy data is often not available.
 Information on electricity load profiles can be confidential and competition-sensitive
 Installing energy monitoring equipment can be expensive
“The level of energy data collection in Irish WWTPs is often limited.” (Fitzsimons et al., 2016)

7. /12
Dana Kirchem dana.kirchem@esri.ie
7
Therefore, we use an integrated energy systems model for the wastewater treatment
process.
electricity
system
Pumped
storage
Battery
storage
Blower
Aerated tank
• Micro organisms
• X = (inflow-outflow) + growth - decay
• Substrate
• S = (inflow-outflow) - consumption
• Dissolved oxygen
DO = (inflow-outflow) + oxygen supply
- consumption
Three-component wastewater process modelPower system model

8. /12
Dana Kirchem dana.kirchem@esri.ie
8
These models come from two completely different classes of models.
Energy system models Wastewater treatment plant models
Mathematical approach (MI)LP System of non-linear differential
equations
Subject of analysis System operation (supply side) Biochemical processes
Integration End-users as black-boxes Closed system
Model elements Objective function and system
constraints
Process rates and mass balance
equations
Aim of model System cost minimisation Evaluation of control strategies
Time frame Discrete time steps Continuous time

9. /12
Dana Kirchem dana.kirchem@esri.ie
Linear model
Locally valid
Weighted linear
combinations,
following Smets et al.
(2003)
9
The integration requires a reduction and linearization of the standard
wastewater treatment model.
Standard model ASM1
(Activated sludge model no. 1)
heterotrophic biomass BH
autotrophic biomass BA
readily biodegradable substrate SS
slowly biodegradable substrate XS
inert particulates XP
particulate organic nitrogen XND
soluble organic nitrogen SND
ammonia SNH
nitrate SNO
dissolved oxygen SO
Reduced model
biomass X
Substrate S
𝛼 and 𝛽
dissolved oxygen DO
10000
20000
30000
m3/h
Wastewater inflow categories
low
mid
high
Linear model
Globally valid
Switching function with
binary variables

10. /12
Dana Kirchem dana.kirchem@esri.ie
10
The reduced model yields reasonable results for the DO consumption and Biological
oxygen demand (BOD5).
TKN = Total Kjeldahl
Nitrogen

11. /12
Dana Kirchem dana.kirchem@esri.ie
11
The new framework will help to unlock industrial demand response more effectively.
 More accurate results for demand response activity and its system effects.
 Taking into account process constraints (e.g. effluent standards in wastewater
treatment)
 Overcoming energy data limitations
 The framework can be applied to other industrial processes
 Nonlinearities occur in many industries

12. /12
Dana Kirchem dana.kirchem@esri.ie
12
Acknowledgements
This publication has emanated from research conducted with the financial support of Science Foundation Ireland
under the SFI Strategic Partnership Programme Grant Number SFI/15/SPP/E3125. The opinions, findings and
conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect
the views of the Science Foundation Ireland.
Thank you very much for your
attention!